Fetch prices

# Look for files of interest
files = list.files(path = "tmp", pattern = "*.json", full.names = TRUE)
stopifnot(length(files) > 0)
print(files)

##  [1] "tmp/BCH.json"    "tmp/BCHABC.json" "tmp/BTC.json"   
##  [4] "tmp/EOS.json"    "tmp/ETC.json"    "tmp/ETH.json"   
##  [7] "tmp/LTC.json"    "tmp/REP.json"    "tmp/XLM.json"   
## [10] "tmp/XRP.json"    "tmp/ZRX.json"

# Get the times from the first coin
raw = fromJSON(file = files[1])
timestamp = c(); for (observation in raw$Data) timestamp = c(timestamp, observation$time)

# Fix time offset (Epoch to friendly units)
timestamp = timestamp - timestamp[1]
timestamp = timestamp / 3600

prices <- cbind(timestamp)

# Add a new column for each coin
for (f in files) {
  raw = fromJSON(file = f)
  p = c(); for (observation in raw$Data) p = c(p, observation$close)
  prices <- cbind(prices, usd = p)
}

Linear regression

# X-axis is time
x <- prices[,1]

# Plot prices on the Y-axis
for (i in 3:12) {

  y <- prices[,i]
  relation <- lm(y ~ x)

  par(bg = "gray94")
  plot(x, y, col = "orange", main = files[i - 1], 
       abline(relation, col="red", lty = 10), xlab = "hours", ylab = "USD")
}

Normalise

# Take a copy to normalise
n = prices

for (i in 2:12) {
  low = min(n[,i])
  high = max(n[,i])

  n[,i] = n[,i] - low
  n[,i] = n[,i] / (high - low)
}

matplot(n[,2:12], pch = 1)

Correlation

for (i in 3:12) {
  cor(prices[,2], prices[,i])
  plot(prices[,2], prices[,i])
}